SERS-Based Label-Free Insulin Detection at ... - ACS Publications

functions in their natural state and expedite diabetes-related clinical ..... the substrate to collect and average more Raman-signal data points as op...
0 downloads 0 Views 1MB Size
Subscriber access provided by READING UNIV

Letter

SERS-Based Label-Free Insulin Detection at Physiological Concentrations for Analysis of Islet Performance Hyunjun Cho, Shailabh Kumar, Daejong Yang, Sagar R. Vaidyanathan, Kelly Woo, Ian Garcia, Hao Jan Shue, Youngzoon Yoon, Kevin Ferreri, and Hyuck Choo ACS Sens., Just Accepted Manuscript • DOI: 10.1021/acssensors.7b00864 • Publication Date (Web): 11 Jan 2018 Downloaded from http://pubs.acs.org on January 12, 2018

Just Accepted “Just Accepted” manuscripts have been peer-reviewed and accepted for publication. They are posted online prior to technical editing, formatting for publication and author proofing. The American Chemical Society provides “Just Accepted” as a free service to the research community to expedite the dissemination of scientific material as soon as possible after acceptance. “Just Accepted” manuscripts appear in full in PDF format accompanied by an HTML abstract. “Just Accepted” manuscripts have been fully peer reviewed, but should not be considered the official version of record. They are accessible to all readers and citable by the Digital Object Identifier (DOI®). “Just Accepted” is an optional service offered to authors. Therefore, the “Just Accepted” Web site may not include all articles that will be published in the journal. After a manuscript is technically edited and formatted, it will be removed from the “Just Accepted” Web site and published as an ASAP article. Note that technical editing may introduce minor changes to the manuscript text and/or graphics which could affect content, and all legal disclaimers and ethical guidelines that apply to the journal pertain. ACS cannot be held responsible for errors or consequences arising from the use of information contained in these “Just Accepted” manuscripts.

ACS Sensors is published by the American Chemical Society. 1155 Sixteenth Street N.W., Washington, DC 20036 Published by American Chemical Society. Copyright © American Chemical Society. However, no copyright claim is made to original U.S. Government works, or works produced by employees of any Commonwealth realm Crown government in the course of their duties.

Page 1 of 23 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

ACS Sensors

SERS-Based Label-Free Insulin Detection at Physiological Concentrations for Analysis of Islet Performance Hyunjun Cho1,†, Shailabh Kumar2,†, Daejong Yang2, Sagar Vaidyanathan1, Kelly Woo1, Ian Garcia1, Hao J. Shue1, Youngzoon Yoon3, Kevin Ferreri4, Hyuck Choo*,1,2

1

Department of Electrical Engineering, 2Department of Medical Engineering, California Institute of Technology, Pasadena, CA 91125, United States

3

Device Lab, Device & System Research Center, Samsung Advanced Institute of Technology (SAIT), Suwon, 16678, Republic of Korea

4

Department of Translational Research and Cellular Therapeutics, Diabetes and Metabolism Research Institute, Beckman Research Institute of the City of Hope, Duarte, CA

† These authors contributed equally Corresponding author *E-mail: [email protected]

1 ACS Paragon Plus Environment

ACS Sensors 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 2 of 23

Abstract Label-free optical detection of insulin would allow in vitro assessment of pancreatic cell functions in their natural state and expedite diabetes-related clinical research and treatment, however no existing method has met these criteria at physiological concentrations. Using spatially-uniform 3D gold-nanoparticle sensors, we have demonstrated surface-enhanced Raman sensing of insulin in the secretions from human pancreatic islets under low and high glucose environments without the use of labels such as antibodies or aptamers. Label-free measurements of the islet secretions showed excellent correlation among the ambient glucose levels, secreted insulin concentrations, and measured Raman-emission intensities. When excited at 785 nm, plasmonic hotspots of the densely-arranged 3D gold-nanoparticle pillars as well as strong interaction between sulphide linkages of the insulin molecules and the gold nanoparticles produced highly sensitive and reliable insulin measurements down to 100 pM. The sensors exhibited a dynamic range of 100 pM to 50 nM with an estimated detection limit of 35 pM, which covers the reported concentration range of insulin observed in pancreatic cell secretions. The sensitivity of this approach is approximately four orders of magnitude greater than previously reported results using label-free optical approaches, and it is much more costeffective than immunoassay-based insulin detection widely used in clinics and laboratories. These promising results may open up new opportunities for insulin sensing in research and clinical applications.

Keywords: SERS, Insulin, biosensing, plasmonics, nanoparticles

2 ACS Paragon Plus Environment

Page 3 of 23 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

ACS Sensors

Hormones are chemical messengers that control a wide variety of functions in the human body. Maintaining adequate hormone levels is extremely important for human health and disruption to these levels can result in life-debilitating conditions. Simple and easy measurements of hormonal secretions ex vivo or in vivo are essential for implementing next generation biosensors, allowing convenient monitoring of health and early disease detection. One of the most prevalent diseases resulting from hormonal dysfunction is diabetes, which arises from a disruption in the release of insulin in the body.1-2 Insulin is a peptide hormone which is secreted by beta cells, one of five primary cell-types which populate pancreatic cellular clusters known as islets. The concentration of insulin secreted from beta cells in plasma has been reported to vary between 100 pM (fasting) and 2 nM (about 1 hour after glucose intake) in non-diabetic individuals.3-4 In diabetic individuals, functional damage to the beta cells reduces or inhibits their ability to release insulin. One of the leading methodologies for treatment of type-1 diabetes is pancreatic islet transplantation, where healthy islets harvested from deceased donors are transplanted into diabetic patients.5-6 Since the number of donors is limited, methodologies that can improve the efficiency and success of the transplantation process are urgently needed. Before transplantation these islets are screened for their viability and functional response to changing glucose concentration in order to reduce the chances of transplant failure.7-9 Sensors capable of detecting secretion of insulin from beta cells in a costeffective, label-free manner with minimal stress to the cells can serve vital roles in clinical quality assessment of islets. External monitoring of insulin concentration can also help in disease diagnosis and management as well as prevent potentially fatal insulin overdoses or hypoglycemia in diabetic patients.10

3 ACS Paragon Plus Environment

ACS Sensors 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 4 of 23

Various sensing methods have been previously explored for insulin detection including radioimmunoassays,11 mass-spectrometry,12 photoluminescence,13 electrochemical methods,14 electrophoresis-dependent

immunoassays,15-16

surface

plasmon-resonance

(SPR)-based

competitive binding assays,17 and fluorescence resonance energy transfer (FRET).18 The optical detection techniques such as SPR, photoluminescence, and FRET suffer from poor sensitivity and cannot detect insulin at physiologically observed picomolar concentrations.13,

17-18

Electrochemical impedance spectroscopy and immunoassays have reported more sensitive insulin-specific detection, however these methods have primarily relied on capture agents such as antibodies for detection.11, 14, 19 These antibodies are not only expensive, but have also been shown to disrupt the natural behavior of live cells.20-21 For applications such as pancreatic islet transplantation, a label-free, optical sensing method capable of performing ultrasensitive detection of insulin without using capturing agents can help improve the cost-effectiveness as well help the islet cells remain as close to their natural state as possible during the pre-surgical screening process. Surface-enhanced Raman spectroscopy (SERS) is an ideal approach for optical label-free sensing because it identifies targeted molecules based on their unique vibrational and rotational signatures.22-26 Application of SERS for hormone detection appears relatively unexplored due to minimal experimental success: previously reported SERS-based quantitative insulin sensors were limited due to weaklyenhancing substrates made of randomly dispersed nanoparticles resulting in micromolar detection sensitivity, approximately 2 to 4 orders of magnitude larger than clinically-relevant insulin levels.27-30 In this study, we report highly sensitive SERS-based insulin sensing at clinically relevant concentrations using a non-resonant SERS substrate with strong signal enhancement and wafer4 ACS Paragon Plus Environment

Page 5 of 23 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

ACS Sensors

scale uniformity.31-34 The sensors were batch-fabricated using a simple two-step process that produced 3 dimensional (3D) gold-nanoparticle (Au-NP) clusters packed densely in a vertical, pillar-like arrangement.33 The 3D pillar-like geometry of the clusters adsorb and excite insulin molecules with vertically packed plasmonic hotspots uniformly over a large surface area for enhanced Raman emissions. This geometry provides significantly stronger and more spatially uniform enhancement than previously used substrates and greatly improved the detection limit.3136

The obtained result is several orders of magnitude better than previously reported values

produced using SERS approaches,27-30 and the detection limit of the substrate is estimated as 35 pM. Using our method, we characterized the insulin concentrations in pancreatic islet secretions that were collected under low and high glucose conditions. The measurements obtained using our approach showed excellent consistency and paralleled the concentrations obtained using an enzyme-linked immunosorbent assay (ELISA). The fabrication of the SERS substrate was described in detail by Yang et al.33 A summary of the fabrication has been provided in the Supporting Information (Figure S1). The cross-sectional view of the 3D Au-NP SERS substrate acquired using scanning electron microscopy (SEM) is shown in Figure 1a. The SEM image reveals a dense, vertical arrangement of gold nanoparticle clusters that, based on experimental results, provide plasmonic hotspots necessary for ultrasensitive optical detection.33 SEM images in Figure 1 (b) and (c) show Au-NP clusters before and after adsorption of 1 µM insulin, respectively. Raman spectra obtained from powdered insulin (Figure 1d) and 3D Au-NP clusters incubated in 1 mM insulin (Figure 1e) were collected as reference Raman signatures for insulin sensing and show consistent peak intensities and location. The most intense peak at 1002 cm-1 corresponding to the ring-breathing mode of aromatic phenylalanine was used to monitor the change in Raman intensity as a function of

5 ACS Paragon Plus Environment

ACS Sensors 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 6 of 23

insulin concentration.37 Measurements were performed at the center of each chip and at four additional points -- top, bottom, left, and right, about 0.5 mm away from the center location -- to characterize the Raman intensity and spatial uniformity. The change in SERS signal intensity as a function of insulin concentration between 100 pM to 50 nM is shown in Figures 2 (a) and (b). The 1002 cm-1 peak was identifiable at 100 pM and a linear increase in the Raman intensity was observed as the insulin concentration increased from 100 pM to 10 nM. The Raman intensity reached saturation at insulin concentrations greater than 50 nM, which indicates an almost complete monolayer coverage. Insulin can exist in solutions as a hexamer (~ 5 nm in diameter and 3.5 nm in height) in the presence of zinc ions at neutral pH, as a dimer ( ~ 4.5 nm diameter and ~ 2 nm in height) or as a monomer (~ 2 nm diameter and height ~ 2 nm).38 As no zinc ions were added to insulin containing buffers, the molecules are expected to be in a monomer-dimer equilibrium. Given the size of insulin molecules, even if there were a second layer sitting on top of the first layer, the second layer would be too far away from the surface to significantly benefit from the near-field enhancement on the surface, and the Raman-emission contribution by the second and/or third layers would be minimal.39 The signal-to-noise ratio (SNR) at 100 pM was calculated to be approximately 8.5, and the theoretical detection limit was calculated to be 35 pM for a minimum acceptable SNR of 3. For typical SERS substrates, even a small displacement of the incident laser could often lead to a huge change in Raman signal intensity due to non-uniformly distributed SERS hotspots. In order to improve the consistency of SERS measurements, recent focus has been placed on improving the chip-scale measurement uniformity either by implementing better substrates,31-34 or employing better measurement strategies such as areal averaging,35-36 which involves scanning a larger area on the substrate to collect and average more Raman-signal data points as opposed to

6 ACS Paragon Plus Environment

Page 7 of 23 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

ACS Sensors

making a single point measurement. Two-dimensional (2D) mapping of Raman spectra was performed over a 1×1 mm2 area with a step size of 20 µm on a chip coated with a 10-nM insulin solution (Figure 2c). The relative standard deviation (RSD) of the signal fluctuation (calculated using five points distributed across the substrate as mentioned earlier) was 4.3%, indicating uniform enhancement of Raman signal over a large area of the sensor. A histogram of the Raman signals obtained over the mapped region also reveals a very narrow intensity distribution and indicates uniform enhancement (Figure 2d). Both highly uniform 2D areal scan results (Figure 2c, d) and the saturation of the Raman-intensity curve support that the insulin coverage on the surface is a monolayer – or the major contribution comes from the monolayer, and the transport mechanism to the surface is a relatively uniform diffusion process. Areal averaging was also explored to check if we can further improve the spatial uniformity of Raman measurements. Figure 2e shows the RSD as a function of the scan area used for averaging. The results show that increasing the averaging scan area improves the measurement uniformity. For example, increasing the average scan area from 50×50 µm2 to 400×400 µm2 reduces the RSD from 3.6% to 1.7%. However, it should be noted the improvement in measurement uniformity from areal averaging is small, only about 1-3%, which is a testament to the excellent spatial uniformity of signal obtained from the 3D Au-NP substrate. This excellent spatial uniformity of SERS measurements also indicates a complete and uniform coverage of insulin on the substrate. In order to demonstrate a practical, rapid detection of low-level insulin using the SERS substrate, a 200 µL drop of 100-pM insulin solution was applied to the surface of fresh SERS chips and dried by placing on a hotplate (without stirring) at 50˚C for 30 minutes. The locations and intensities of the Raman peaks measured from the evaporation-prepared samples (Figure 2f) matched the substrates incubated in insulin solutions for 12 hours. The droplet is typically 0.5 cm

7 ACS Paragon Plus Environment

ACS Sensors 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 8 of 23

in diameter, and signal was collected from central regions of the placed droplet, where insulin transport should be governed by diffusion and any thermal gradient-derived convective currents, thus avoiding the edges where evaporation-driven concentrating ring effects are observed. As mentioned earlier, even if more than one layer of molecules were assembled on the surface, the second layer would be too far away from the surface to benefit from the near-field enhancement on the surface as the Raman-emission contribution by the second and/or third layers would be minimal.39 A